mcp-defense is a CLI tool for AI infrastructure engineers and security teams, providing kernel-native runtime defense for AI agents. It enforces MCP tool-call policies and correlates kernel events to secure agentic workloads such as Claude Code, Codex, and Devin.
In the Security & compliance platforms space, mcp-defense takes a focused approach. It focuses on securing AI agent workloads at runtime by enforcing kernel-level policies and monitoring tool calls. mcp-defense is a B2B product aimed at AI infrastructure engineers and security teams. mcp-defense is paid. It runs on the command line.
jessfortemnaturae8717 builds and maintains mcp-defense, and it first shipped in 2026. Key capabilities include runtime defense, kernel correlation, and MCP policy enforcement. It exposes integrations via an MCP server.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do